WO1991007506A1

WO1991007506A1 - A method of synthesizing double-stranded dna molecules

Info

Publication number: WO1991007506A1
Application number: PCT/US1990/006557
Authority: WO
Inventors: Douglas Howard Jones
Original assignee: The United States Of America, Represented By The Secretary, United States Department Of Commerce
Priority date: 1989-11-08
Filing date: 1990-11-08
Publication date: 1991-05-30
Also published as: AU6886991A

Abstract

The present invention relates to a method in which the polymerase chain reaction (PCR) is used to generate products that form double-stranded DNA with discrete, cohesive, single-stranded ends when said products are combined, denatured, and reannealed. Such cohesive ends can be used to form DNA joints, as well as circular DNA. Furthermore, the above method can be utilized in site-specific mutagenesis, in DNA recombination, and in other additional techniques.

Description

A METHOD OF SYNTHESIZING DOUBLE-STRANDED DNA MOLECULES

BACKGROUND OF THE INVENTION Technical Field The subject invention relates to a method in which the polymerase chain reaction (PCR) is used to generate products that when combined, denatured, and reannealed, form double-stranded DNA with discrete, cohesive, single- stranded ends. Such cohesive ends can be utilized in the formation of precise, directional DNA joints without dependence on enzyme restriction sites.

Furthermore, the placement of cohesive single strands on both ends of a DNA molecule permits a DNA strand or strands to be annealed to form circular DNA. These recombinant circles of DNA can be directly trans- fected into appropriate host cells. Moreover, since no enzymatic step is needed apart from the PCR amplification, the method of generating recombinant circles upon rean- nealing PCR generated products simplifies DNA utagenesis, recombination, and cloning. Background Information

The polymerase chain reaction (PCR) is a method by which a specific DNA sequence can be amplified in vitro

(U.S. Patent 4,683,202, U.S. patent 4,683,195, Mullis et al, Cold Spring Harbor Symposia on Quantitative Biology Vol. LI: 263-73 (1986), 3aiki et ^~ . , Science 239:487-91

(1988) and Saiki et ai. Science 230:1350-54 (1985)). Frit- investigato s have used PCR to generate site-specif¬ ic ii tants (Hemsley et al, -Nucleic Acids Res. 17:6545-51 ( 1989 ) , Higuchi et . al', Nucleic Acids Res. 16:7535-67 (1588), Ho et alv. Gene 77:51-59 (1989), Kadowaki, Gene 76:161-66 (1989), Kammann et al, Nucleic Acids Res. 17:5404 (1989), Nelson et al, Anal. Biochem. 180:147-51

(1989) and Vallette et al., Nucleic Acids Res. 17:723-33 (1989). PCR has also been used to amplify inserts which later undergo a separate subcloning procedure (Saiki et al, Science 239:487-91 (1988)). . Site-specific mutants are created by introducing mismatches nto the oligonucleotides used to prime the PCR amplification. These oligonucleotides, with their mutant sequence, are incorporated into the PCR product. PCR has also been used to join segments of DNA by a method called splicing by overlap extension (Higuchi et al, Nucleic Acids Res. 16:7351-67 (1988), Ho et al, Gene 77:51-59 (1989), Horton et al, Gene 77:61-68 (1989)). This re¬ quires two sequential PCR amplifications, and results in a blunt-ended product. Since PCR generated products are blunt-ended, a cloning step into a vector is necessary before individual clones can be replicated in plasmids and evaluated. This cloning step, to date, has necessitated restriction enzyme digestion of the PCR product and has always required a ligation step.

A recent report describes a method for site- specific mutagenesis based on amplification of the entire plasmid (Hemsley et al, Nucleic Acids Res. 17:6545-51 (1989)). In that protocol, the ends of the PCR product are treated with the Klenow fragment of DNA polymerase 1. Subsequently, these ends are phosphorylated at the 5' terminus with Polynucleotide Kinase prior to an in vitro self-annealing blunt end ligation reaction.

SUMMARY OF THE INVENTION In the method of the present invention, DNA joints are formed by the use of PCR amplifications wherein the products of these amplifications combine, denature and reanneal so as to form double-stranded DNA with cohesive single-stranded nds. Such ends anneal to form the DNA joints. , ^■ .furthermore, the present method also generates

- recombinant DNA circles, without the utilization of an enzyme digestion or ligation. eaction. The circles of DNA can be transfected directly into an appropriate host where ligation and replication occur. Thus, the method by which such circles are formed, referred to as "recombinant circle PCR" (RCPCR) , greatly simplifies DNA mutagenesis, recombination, and cloning. The rapid execution of complex vector reconstructions involving deletions, insertions and substitutions without regard to restriction enzyme sites, and without the use of linkers or adaptors, can also be accomplished using this method. Furthermore, this technique requires relatively few cycles of PCR amplification to obtain the clone of interest, thereby decreasing the occurrence of unwanted mutations. Such mutations occurring during PCR amplification have been noted in several studies. (Dunning et al, Nucleic Acids Res. 16:10393 (1988), Fucharoen et al, J. Biol. Chem. 264:7780-83 (1989), Saiki et al, Science 239:487-491 (1988) and Tindall et al, Biochemistry 27:6008-13 (1988)). The most basic aspect of the present invention relates to a method for synthesis of a double-stranded DNA molecule using the polymerase chain reaction process, wherein the nucleotide sequence of the first single-strand of said double-stranded DNA molecule has at least one difference from the complement of the nucleotide sequence of the second single-strand of said double-stranded DNA molecule, comprising the steps of:

(i) providing a single-stranded DNA template, any desired segment of the nucleotide sequence of said tem- plate being used for synthesis of said first strand of said double-stranded DNA molecule; and

(ii) providing a first polynucleotide primer, at least a portion of the nucleotide sequence of which is complementary to said template starting at the nucleotide that defines the 3' end of said desired segment and extending ir the 5' direction of said template; and

(iii) providing a second polynucleotide primer, at least a portion of the nucleotide sequence of which is cnmr^ιementary to saxd template , starting atr^'said 3' end of said desired segment or star ng 5' to"said 3' end, and extending in the 5' direction z said template, wherein there is at 1 :st one difference between said nucleotide sequence of »aid first primer and said nucleotide sequence of said second primer, said difference comprising a substitution , deletion, insertion or exten¬ sion of one or more nucleotides, provided that, if both of said first and second primers are complementary to said template starting at the same nucleotide on said 3' end of said desired segment, and if said difference in said first and second primers is entirely 5' to said 3' end of said desired segment, then said difference does not consist entirely of an extension of the sequence that is complementary to said template in one of said primers with respect to the other; and

(iv) providing a third polynucleotide primer, at least a portion of the nucleotide sequence of which is homologous to said single-stranded template starting at the nucleotide that defines the 5' end of said desired segment and extending in the 3' direction of said tem¬ plate; and

(v) synthesizing double-stranded DNA molecules using said template and said primers and the polymerase chain reaction process; and

(vi) separating the primer extension products resulting from step (v) from the strands to which they are hybridized so that single-stranded extension products are formed; and (vii) reannealing the single strands resulting from step (vi) .

Another aspect of the present invention relates to a method for producing a circular double-stranded DNA molecule in which steps (v), (vi), and (vii) above further comprise the steps of: in a first container means containing a first aliquot of a DNA template:

(i) treating a first strand of said template with a first primer having a sequence that is sufficiently complementary to a first region of said first strand to hybridize therewith under conditions such that hybridiza¬ tion between said first primer and said first region of said first strand occurs;

(ii) treating a second strand of said template with a second primer having a sequence that is sufficiently complementary to a first region of said second strand to hybridize therewith under conditions such that hybridization between said second primer and said first region of said second strand occurs;

(iii) treating the products of steps (i) and

(ii) using denaturation and primer reannealing such that an extension product of each of said first and second primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said first and second primers, when separated from the strand to which it is hybridized, can serve as a template for synthesis of the extension product of the other of said first and second primers;

(iv) separating the primer extension prod¬ ucts resulting from step (iii) from the strands to which they are hybridized so that single-stranded extension products are formed; and

(v) treating the single-stranded extension products from step (iv) with said first and second primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (iv) as a template, whereby a first double-stranded product is formed; in a second container means containing a second aliquot of said template:

(vi) treating said second strand of said template with a third primer having a sequence that is su-.^"iciently complementary to a second region of said sec nd strand to hybridize therewith under conditions such that hybridization between said third primer and said se-^ d rpgion of said second strand occurs, said third primer having a sequence that is sufficiently cpmplementa- ry to the sequence of said first primer for hybridization to occur;

(vii) treating said first strand of said plasmid with a fourth primer having a sequence that is sufficiently complementary to a second region of such first strand to hybridize therewith under conditions such that hybridization between said second region of said first strand and said fourth primer occurs, said second region of said first strand being distinct from said first region of said second strand;

(viii) treating the products of steps (vi) and (vii) under conditions such that an extension product of each of said third and fourth primers is synthesized which is complementary to each of said second and first strands, respectively, such that the extension product synthesized from one of said third and fourth primers, when separated from the strand to which it is hybridized, can serve as a template for synthesis of the extension product of the other of said third and fourth primers;

(ix) separating the primer extension prod¬ ucts resulting from step (viii) from the strand to which it is hybridized so that single-stranded extension prod- ucts are formed; and

(x) treating the single-stranded extension products from step (ix) with said third and fourth primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (ix) as a template whereby a second double-stranded product is formed;

(xi) combining the first and second double- stranded products resulting from steps (v) and (x) ;

(xii) separating the strands of said first and second double-stranded products resulting from steps (v) and (x) ;

(xiii) reannealing the single strands result¬ ing from step (xii); and

(xiv) .selecting circular DNA molecules comprising the extension products of said first, second, third and fourth primers.

An additional aspect of the method shown directly above involves the situation wherein said first and third primers contain at least one base that is not complementa- ry to said first region of said first strand of said template and said second region of said second strand of said template, respectively, and wherein said at least one base of said first primer is complementary with said at least one base of said third primer.

Furthermore, another aspect of the present inven¬ tion utilizing the method relating to synthesis of a double-stranded DNA molecule using the PCR process in- volves transferring an insert contained in a donor plasmid to a recipient plasmid in which (v) , (vi), and (vii) further comprise the steps of: in a first container means containing an aliquot of a donor plasmid containing an insert: (i) treating a first strand of said donor plasmid with a first primer having a sequence that is sufficiently complementary to a region of said first strand to hybridize therewith under conditions such that hybridization between said first primer and said region of said first strand occurs, wherein the first primer con¬ tains only a sequence present in one end of the insert;

(ii) treating a second strand of said donor plasmid with a second primer having a sequence that is sufficiently complementary to a region of said second strand to hybridize therewith under conditions such that hybridization between said second primer and said region of said second strand occurs, wherein the second primer contains only a sequence present in the other end of the insert; in a second container means containing an aliquot of a donor plasmid containing an insert:

(iii) treating said first strand of said donor plasmid with a third primer having a sequence that is sufficxently complementary to said-region of said first stranr to hybridize therewith under conditions such that hybri< ;ation between said third primer and said region of said rst strand occurs, wherein the 3' end of said third prime: is identical to primer 1 and has several base pairs on its 5' end which are homologous to a region on the recipient plasmid of step (xi);

(iv) treating said second strand of said donor plasmid with a fourth primer having a sequence that is sufficiently complementary to said region of said second strand to hybridize therewith under conditions such that hybridization between said fourth primer and said region of said second strand occurs, wherein the 3' end of said fourth primer is identical to primer 2 and has several base pairs on its 5' end which are homologous to a region on the recipient plasmid of step (xi);

(v) treating the products of steps (i) and (ii) under conditions such that an extension product of each of said first and second primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said first and second primers, when separated from the strand to which it is hybridized can serve as a template for synthesis of the extension product of the other of said first and second primers;

(vi) separating the primer extension prod¬ ucts resulting from step (v) from the strands to which they are hybridized so that single-stranded extension products are formed; and (vϋ) treating the single-stranded extension products from step (vi) with said first and second primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (vi) as a template, whereby a first double-stranded product is formed;

(viii) treating the products of steps (iii) and (iv) under conditions such that an extension product of each of said third and fourth primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension, product synthesized from one of said third and fourth primers, when separated from the strand to which it is hybridized, can serve as a template for synthesis of the extension product of the other of said third and fourth primers; (ix) separating the primer extension prod¬ ucts resulting from step (viii) from the strands to which they are hybridized so that single-stranded extension products are formed; and (x) treating the single-stranded extension products from step (ix) with said third and fourth primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (ix) as a template, whereby a first double-stranded product is formed; in a third container means containing an aliquot of a recipient plasmid:

(xi) treating the first strand of said recipient plasmid with a sixth primer having a sequence that is sufficiently complementary to a first region of said first strand to hybridize therewith under conditions such that hybridization between said sixth primer and said first region of said first strand occurs, said sixth primer containing the region complementary to the 5' regions of primer 3;

(xii) treating the second strand of said recipient plasmid with a fifth primer having a sequence that is sufficiently complementary to a first region of said second strand to hybridize therewith under conditions such that hybridization between said first region of said second strand and said fifth primer occurs, said fifth primer containing the region complementary to the 5' region of primer 4; in a fourth container means containing an aliquot of a recipient plasmid:

(xiii) treating the first strand of said recipient plasmid with an eighth primer having a sequence that is sufficiently complementary to a second region of said first strand to hybridize therewith under conditions such that hybridization between said eighth primer and said second region of said first strand occurs, said second region of said first strand being distinct from said first region of said first strand; (xiv) treating the second strand of said recipient plasmid with a seventh primer that is suffi¬ ciently complementary to a second region of said second strand to hybridize therewith under conditions such that hybridization between said seventh primer and said second region of said second strand occurs, said second region of said second strand being distinct from said first region of said second strand; (^χv) treating the products of steps (xi) and

(xii) under conditions such that an extension product of each of said sixth and fifth primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said sixth and fifth primers, when separated from the strand to which it is hybridized, can serve as a template for the synthesis of the extension product of the other said sixth and fifth primers;

(xvi) separating the primer extension prod- ucts resulting from step (xv) from the strands to which they are hybridized so that single-stranded extension products are formed; and

(xvii) treating the single-stranded exten¬ sion products from step (xvi) with said sixth and fifth primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (xvi) as a template, whereby a first double-stranded product is formed;

(xviii) treating the products of steps (xiii) and (xiv) under conditions such that an extension product of each of said eighth and seventh primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said eighth and seventh primers, when separated from the strand to which it is hybridized, can serve as a template for synthesis of the extension product of the other of said eighth and seventh primers;

(xix) separating the primer extension products resulting from step (xviii) from the strands to which they are hybridized so that single-stranded exten¬ sion products are formed; and

(xx) treating the single-stranded extension products from step (xix) with said eighth and seventh primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (xix) as a template, whereby a first double-stranded product is formed;

(xxi) combining the double-stranded prod¬ ucts resulting from steps (vii), (x) , (xvii) and (xx);

(xxii) separating the strands of said double-stranded products resulting from steps (vii), (x), (xvii) and (xx); and

(xxiii) reannealing the single-strands resulting from step (xxii).

An additional aspect of the present invention involves a method of producing a site-specific mutation in an autonomously replicating double-stranded circular DNA comprising the steps of: in a first container means containing a first aliquot of said circular DNA:

(i) contacting said circular DNA with a first primer and second primer, wherein said first primer is complementary to a region of the first strand of said circular DNA, except that it contains a mismatched base at the site at which mutagenesis is to end of said first primer, said second primer is complementary to the second strand of said circular DNA, wherein the 5' ends of said first and second primers bind at contiguous but n~ - overlapping regions of said first and second strands, nd wherein sa d primers are ir subs ^tantial molar excess over said strands, (ii) producing non-circular copies of said second strand containing said mismatched based and non- circular copies of said first strand by means of the polymerase chain reaction, in a second container means containing a second aliquot of said circular DNA:

(iii) contacting said circular DNA with a third primer and a fourth primer, wherein said third primer is complementary to a region of said second strand, except that it contains a mismatched base at the site at which mutagenesis is to occur, said mismatched base being at least 8 bases from the 5' end of said third primer, said third primer containing a sequence including said mismatched base which is complementary to a fourth primer which is complementary to said first strand, wherein the 5' ends of said third and fourth primers bind at contigu¬ ous but non-overlapping regions of said first and second strands, and wherein said primers are in substantial molar excess over said strands,

(iv) producing non-circular copies of said first strand containing said mismatched base and non- circular copies of said second strand by means of the polymerase chain reaction, (v) combining the products of steps (ii) and

(iv),

(vi) subjecting the mixture of step (v) to denaturation conditions to produce single-stranded DNA,

(vii) subjecting the mixture of step (vi) to annealing conditions to produce double-stranded DNA, and

(viii) isolating the circular DNA containing the mismatched bases.

Moreover, the autonomously replicating double- stranded circular DNA of the above method can be a plas- mid.

This basic method of the present invention•-forms the basis of a variety of embodiments that are designed to accomplish various genetic engineering objectives .requir¬ ing'cutting and splicing of DNA strands. These operations - -a-re presently performed using restriction enzymes and ligases. For example, the present method may be used to create cohesive single-stranded ends useful for joining double-stranded DNA molecules. These cohesive ends can be made sufficiently long to enable direct introduction of the joined molecules into a host cell, a bacterial cell, for example, which can covalently close the joint by DNA repair mechanisms without the need for prior extracellular ligation. Therefore, for many routine recombinant DNA applications, the present method obviates the need for restriction enzymes and ligase in routine genetic engi¬ neering of DNA molecules.

In the present method, the polymerase chain reaction process is carried out in the ordinary manner that is known in the art, according to standard methods described hereinbelow, excepting that for synthesis of a double-stranded DNA molecule in which the nucleotide sequence of the first single-strand has at least one difference from the nucleotide sequence of the complement of the second single-strand, the primers are provided according to the steps of the method of this invention as recited above and in the claims.

The template DNA is any single-stranded DNA that is suitable for use as a template in the ordinary poly¬ merase chain reaction process. Thus, the desired segment of this template must be long enough to encompass a sequence that anneals stably to a complementary polynucle- otide primer (i.e., at least about 10 nucleotides long) as well as any additional desired template sequences. The practical upper size limit for efficient amplification of the desired template segment is the same as that known in the art for the polymerase chain reaction process in general. The products of the usual polymerase chain reac¬ tion process using two primers, one complementary to the desired template and the other homologous to that same template, are dc ^~ le~εbranded DNA molecules in which the nucleotide seque. e o_^ the first strand is essentially identical to the complement of the nucleotide sequence of tne second strand. The final product of the present method, in contrast, is a double-stranded DNA molecule in which the nucleotide sequence of the first strand has at least one difference from the complement of the nucleotide sequence of the second strand. In the present method, it is the provision of two different primers that are at least in part complementary to the template that results in the polymerase chain reaction process producing two different double-stranded DNA products: one product containing strands extended from the first primer and a second product containing strands from the second primer. In each of these double-stranded products, however, the sequences of the two strands will be exactly complementa¬ ry.

The object here is to generate from these two different double-stranded DNA molecules, the final product of the present method, namely a double-stranded DNA molecule in which the nucleotide sequence of the first strand has at least one difference from the complement of the nucleotide sequence of the second strand. To attain this object, the strands of these two different double- stranded DNA molecules must be separated (for example, by heating above the melting point for these duplexes) and the resulting single strands from the two double-stranded DNAs must be contacted with each other under conditions such that complementary DNA strands anneal and form double-stranded DNA molecules again. The final product of the present method, a double-stranded DNA molecule in which the nucleotide sequence of the first strand has at least one difference from the complement of the nucleotide sequence of the second strand, comprises one or more of the several forms of double-stranded DNA molecules result- ing from random reassortment of single strands during anneal g and can be selected from this mixture by physi¬ cal or biological means that are well known in the art of genetic enginee-Ting.

While creation of a double-stranded DNA molecule with at least one end having a single-stranded extension is a frequent application of this basic method of this ■invention, the difference between the first strand and the complement of the second strand of the double-stranded DNA molecule of this invention comprises any substitution, deletion, insertion or extension of one or more nucleo- tides of one strand of the double-stranded DNA with respect to the other. The nature and location of each such difference is determined by the nature and locations of differences in the polynucleotide primers used in this method.

The primers of the present invention have all the physical attributes of primers used in the ordinary polymerase chain reaction process. Thus, each primer comprises a sequence of at least about 10 nucleotides that is complementary to its intended template DNA. In this complementary sequence, the two nucleotides at the 3' end of the primer, where new DNA is to be added, are exactly complementary to its template, while one or more of the other nucleotides in the primer sequence that is comple¬ mentary to its template may not be exactly complementary, provided the overall complementarity is sufficient to provide stable annealing of the primer to the template. This aspect of primer design is well known in the art.

Further, the primers of the present invention comprise DNA or RNA polynucleotides including the natural¬ ly occurring nucleotides, found in DNAs or RNAs, or synthetic nucleotides having modified structures (e.g., methylation, sulfur instead of phosphorus, or other changes) that enhance some factor such as stability, for example, without interfering with performance of the polynucleotide as a primer in the polymerase chain reac¬ tion process. In the present method, the first and second primers, singly or together, optionally comprise addition¬ al nucleotide sequences that are not complementary to the desired tem; ate. Si. rlarly- the third primer, either alone ox together with either one or both -of -the first and second primers of this method, optionally comprises additional nucleotide sequences that are not homologous to the desired template. ^' It will be understood by one familiar with the practice of the polymerase chain reac¬ tion process that, for a polynucleotide to function as a primer on any template, such optional sequences that are not complementary to that template are located on the 5' end of the primer (since their presence on the 3' termi¬ nus would block addition of DNA sequences to the primer) . In either the first, second or third primers of the present method, these optional sequences may comprise, for example, a sequence that is complementary to the known sequence of a single-stranded end of some first double- stranded DNA molecule. Application of the present method using this primer sequence that is not complementary to the template provides a second double-stranded DNA mole¬ cule having a single-stranded end that is complementary to the known single-stranded end of the first double-stranded DNA. The length of such optional sequences is at least one nucleotide and is limited only by the practical considerations of preparing and handling intact single- stranded polynucleotide molecules. Thus, for example, to attach a desired DNA segment to a plasmid, a primer of the present invention may include one entire strand of that plasmid, where that entire strand is not complementary to the desired template. Alternatively, as another example, this optional sequence may be used to attach the single- stranded extension of a cleaved restriction enzyme recog- nition site to a double-stranded DNA made from a desired template DNA.

It is known in the art that Taq polymerase that is routinely used in the polymerase chain reaction process frequently extends a newly synthesized DNA strand on its 3' end by about one nucleotide beyond the 5' end of the template, usually by a single adenosine nucleotide. This occurrence contributes to the difficulty of ligating the usual products of the polymerase chain reaction process which, therefore, do not have perfectly matching ends. The present method relates in part to synthesis of double- stranded DNA molecules having at least one single-stranded extension as a result of providing primers according to this method. These extensions of the present invention are either encoded in the template or in the primer and are not the result of the template independent, primer independent errors of the Taq polymerase that are de¬ scribed above. Further, it would be obvious to one of ordinary skill in the art of genetic engineering that, while such errors reduce the yield of the final desired products under certain circumstances, in no case do the consequences of such errors by the polymerase prevent the use of this method in any desired genetic engineering application. For example, where double-stranded DNAs with cohesive single-stranded extensions are created by this method and covalently joined inside a bacterial cell, absent phosphorylation on the 5' ends of the primers, the bacterial host normally removes that part of the DNA containing the superfluous adenosine residues and the repairs the gap to provide the desired double-stranded product without the Taq-generated errors. If cohesive ends made by this process are designed to be ligated in a reaction using only purified ligase, then those cohesive extensions having additional adenosine residues would not be ligated; but even if half the strands produced by the Taq polymerase are not ligatable for this reason, the net reduction in yield of ligated cohesive ends would be a factor of four for each joint (half of the cohesive ends of each of the two double-stranded DNAs to be joined would have the unwanted extensions, only on one strand, those strands that constitute a 3' end in the joint).

The method of the present invention is distin¬ guished from the known method of splicing DNA molecules by overlap extension, described in the background of this application, by the }. ^sent- requirement for two distinct primers, at least a p tion of the nucleotide sequence of ef_ch which s complementary to ■each desired template. In the overlap extension method,' -only a single primer complementary to each template is used.

Inr certain cases, during the application of this basic method of this invention to the synthesis of double- stranded DNA molecules, redundancies in the primers of this method become apparent to one of ordinary skill in genetic engineering. For example, according to the basic method, to generate double-stranded DNA molecules having single-strand extensions on both ends may require the use of six distinct primers. Three primers are needed for each independently created difference (extension, in this case) between the first strand and the complement of the second strand of the final double-stranded DNA product. For each first primer and related second primer used to create one specific extension, in one embodiment of this case, both comprise the same sequence that is complementa¬ ry to the template, and only the first primer comprises additional sequences that are not complementary to the template, these optional sequences providing the desired single-stranded extension.

In this case, if all the appropriate primers and templates and the desired final products according to this basic method are analyzed diagrammatically, it is apparent that under some conditions a single primer may meet the requirements of two or more of the primers of this basic method (e.g., a single primer may be homologous to one template but complementary to another template) . Thus, in the specific case where both the desired template and its complement are to have extensions added onto their 5' ends, for example, to produce a double-stranded DNA with extensions on each end, the basic method specifies that three primers would be required to modify the desired template and three additional primers to modify the complement of that template. However, when all the primers and templates are compared, it is evident that the third primer- for the desired template, which is homologous to that ^template, is also complementary to the other strand which requires primers, i.e., to the complement of the desired template. Accordingly, the ^"third primer foτ the desired template can serve as the first primer for this other strand.

The second primer for the other strand is still required in this instance to make a second modification in a strand of the final double-stranded DNA product. For example, to create the final double-stranded DNA with single-stranded extensions on each end, this second primer for the other strand might also have an optional sequence that is not complementary to its template and that is not contained in the first primer for the other strand. Moreover, the function of the third primer for the other strand in this case is provided by either of the first or second primers for the original desired template. Accord- ingly, creation of a double-stranded DNA molecule with two different single-stranded extensions requires a total of only four primers rather than six, because some of the four primers meet multiple primer requirements.

Another view of the same situation is that multi- pie discrete differences between the first strand and the complement of the second strand in the double-stranded DNA product of the present method can be created by use of a fourth primer which, like the third primer, is at least partly homologous to the desired template, but this fourth primer has at least one difference from the third primer. Further, it will be understood by one of ordinary skill in the art, that in accord with the present invention, for various purposes, it may be advantageous not to combine all the primers and templates in a single reaction. Rather, for example, separate reaction mixtures compris¬ ing, on the one hand, the first and third primers and tem¬ plate, and on the other hand, the second and fourth primers and template, may be prepared and independently subjected to the polymerase chain reaction process. The double-stranded products of the separate reaction mixtures are th^n combined, and their strands are separated and allowec to reassort into the double-stranded DNA molecules comprising the products of the present invention.

Thus, the basic form of the method of this inven- tion, which is described bove, may be applied ±n a single sample that is subjected .o the polymerase chain reaction process or in multijr..e concurrent polymerase chain reaction processes that comprise different desired tem¬ plates and corresponding primers for each different template, to provide complex genetic engineering applica¬ tions, as described in several exemplary embodiments below. Accordingly, in the application of the basic method to selected embodiments of genetic engineering applications below, it will be understood that under the specified conditions, the minimum number of distinct primers required for synthesis of complex DNA constructs will not necessarily be a multiple of the three distinct primers required for the basic method of this invention.

In another aspect, the present invention relates to a particular method of creating DNA joints wherein separate PCR amplifications are used to generate products that when combined, denatured and reannealed, form double- stranded DNA with discrete, cohesive, single-stranded ends. Since these single-stranded ends are designed to be cohesive (i.e., complementary to each other), they will anneal to form DNA joints. The formation of the cohesive ends is a result of the relative placement of primers on the template to be amplified, or the result of the addi¬ tion of a 5' sequence to a primer used to amplify a DNA template.

In a specific application of this technique, the placement of cohesive single strands on both ends of a DNA molecule permits a DNA strand or strands to be annealed to form circular DNA, a method termed recombinant circle PCR (RCPCR) . These recombinant circles of DNA can be directly transfected into appropriate host cells. The method of generating recombinant circles upon reannealing PCR generated products can be used for mutagenesis, recombina¬ tion, and cloning. Ligation of these circular products occurs in the host cells without phosphorylation of primers, restriction enzyme digestion, or ligation in vitro. Therefore, DNA. mutagenesis and recombination can be accomplished without any enzymatic manipulation apart from the PCR amplifications.

The present method possesses several advantages over previous methods used for site-specific mutagenesis and recombination. As noted above, splicing by overlap extension requires two sequential PCR reactions. Since polymerase errors are cumulative during PCR, sequential PCR reactions will result in a greater error rate than the simultaneous PCR reactions of the present method. Furthermore, splic¬ ing by overlap extension generates a blunt ended product that must undergo a conventional subcloning step into a vector. RCPCR generates the mutant and/or recombinant of interest with only one set of PCR reactions, and simulta¬ neously accomplishes the cloning step.

The present invention also offers several advan¬ tages over the method for site-specific mutagenesis described by Hemsley et al., supra. In the latter proto- col, the PCR product is treated with the Klenow fragment of DNA polymerase 1 to even out the ends of the PCR products. Subsequently, these ends are phosphorylated at the 5' terminus with T4 Polynucleotide Kinase, and then must undergo a blunt ended self-annealing reaction with T4 ligase for 8-12 hours prior to transfection. Using this procedure, the desired mutation and reconstitution of the plasmid occurred at a frequency of 82%, and mutant colo¬ nies were obtained in 2-3 days.

In the present method, the recombinant circles of DNA are transfected directly into the appropriate host cells without phosphorylation of the primers or products, and without treatment with Klenow or ligase. In fact, DNA mutagenesis, recombination, and cloning can be accom¬ plished without any enzymatic step apart from the PCR amplifications. The entire RCPCR method, from setting up the PCR reactions to transfection into the appropriate host cells can readily be accomplished in one day, and mutant colonies are obtained within 24-36 hours. Further¬ more, the desired mutation and reconstitution of the plasmid occur at a frequency of 83%—100%.

Additionally, plasmids have been amplified 2.7-3 kb utilizing the present method. Also, an-. bundance of transformants can be generated utilizing only 2 ng of plasmid and using only 14 cycles of amplification. Modification of the PCR conditions of the present method, by increasing initial template amount and/or the number of amplification cycles, for example, should permit the generation of recombinant circle, using larger plasmids. Also, since several distinct and smaller individual segments of large plasmids could be amplified and designed to recombine to form recombinant plasmids, it is possible to mutate and recombine very large plasmids using RCPCR (see Figure 5) .

Another benefit of the present invention relates to nucleotide error frequency. Specifically, since polymerase errors are cumulative during PCR, the nucleo¬ tide error frequency in a PCR generated product increases with the number of cycles. However, RCPCR generates the recombinant of interest with very few PCR cycles, and therefore reduces the risk of acquiring a base error. RCPCR yields the desired product following only 14 cycles of amplification. In contrast, the method described by Hemsley et al. requires 25 cycles of amplification. As noted above, since polymerase errors are cumulative during PCR, the lower number of amplification cycles in the present method will diminish the polymerase error rate. Specifically, using the present method, no errors were detected after sequencing 4100 bases, in the mutagenesis protocol (see Figure 1), indicating a low error frequency of less than 0.025%. In the subcloning protocol (see Figure 2) , only one error was detected after sequencing 900 bases, indicating an error frequency of 0.11%. Basically, RCPCR is a general technique which simplifies and facilitates DNA mutagenesis and recombina¬ tion. Many applications for this method for generating DNA joints, and, in particular, for the use of RCPCR, can be envisioned. In fact, any procedure that involves DNA mutagenesis, recombination, or cloning σan utilize the method of the present invention.

It must be noted that all U.S. patents and all publications referred to herein are hereby incorporated by reference. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates site-specific mutagenesis by the generation of a point mutation using RCPCR. The primers are numbered hemiarrows. The arrows point to the 3' end of each primer and therefore indicate the direction of polymerization. Notches designate point mutations in the primers and resulting mutations in the PCR products. a,b,c, and d are the individual DNA strands produced by the PCR amplification. The "break point" in each amplifi¬ cation reaction is identified by a line drawn on Figure 1. Figure 2 illustrates directional subcloning using RCPCR. The primers are numbered hemiarrows. The DNA strands of the donor plasmid are represented by the thin circles, and the DNA strands of the recipient plasmid are represented with thick circles. The insert in the donor plasmid is cross hatched. Primer 1 and primer 2 contain only sequences contained in the ends of the insert. Primers 3 and 4 have 3' ends that are identical to primers 1 and 2 and have 30 bp 5' ends that are homologous to regions on the recipient plasmid. Primers 5 and 6 amplify the recipient plasmid and contain the regions complementa¬ ry to the 5' segments of primers 3 and 4. Primers 7 and 8 amplify a region of the recipient plasmid which is 30 bp internal to the 5' ends of primers 5 and 6.

Figure 3 illustrates insertional mutagenesis using RCPCR. The primers are numbered hemiarrows. The thick- lined region of primers 1 and 3 are the two complementary sequences of the DNA which are inserted into the plasmid by the RCPCR reaction. a,b,c, and d are the individual DNA strands produced by the PCR amplifications. The heavy-lined regions on the ends of these strands represent the mutating insertional sequence that has been ncorpo¬ rated into the PCR products. Figure 4 illustrates deietional mutagenesis_/ using

CPCR, by the generation of recombinant circles of DNA. .he primers are numbered hemiarrows. The DNA sequence that is deleted is cross-hatched. The 5' end of primer 1 contains a sequence that is homologous to DNA to the right of the region to be deleted, and the 5' end of primer 3 contains a sequence that is homologous to DNA to the left of the region to be deleted. a,b,c, and d are the indi¬ vidual DNA strands produced by the PCR amplifications. The moderately heavy-lined region is DNA that was origi¬ nally to the right of the region that was deleted, and the heavy-lined region is DNA that was originally to the left of the region that was deleted. Figure 5 illustrates site-specific mutagenesis of a very large vector/insert by the generation of recombi¬ nant circles of DNA. The primers are numbered hemiarrows. Notches designate point mutations in the primers and resulting mutations in the PCR products. a,b,c,d,e,f,g, and h are the individual DNA strands produced by the PCR amplifications. The dotted lines identify the regions between adjacent 5' ends of all the primers, thereby making it easier to visualize the formation of the recom¬ binant circles. Figure 6 illustrates substitution of an insert for a sequence of DNA in a recipient plasmid, using RCPCR. The primers are numbered hemiarrows. The DNA strands of the donor plasmid are represented with thick circles. The insert in the donor plasmid is cross-hatched. Primer 1 and primer 2 contain only sequences contained in the ends of the insert. Primers 3 and 4 have 3' ends that are identical to primers 1 and 2 and 5' ends that are homolo¬ gous to the region in which the insertion is to take place on the recipient plasmid. Primers 5 and 6 amplify the recipient plasmid minus the region which is to be removed during the substitution, and contain the regions comple¬ mentary to the 5' regions of primers 3 and 4. Primers 7 end 8 amplify a region of the recipient plasmid which is internal to the 5' ends of primers 5 and 6. . Figure 7 illustrates the joining of one end of a double-stranded DNA molecule to a single-stranded DNA molecule. Primer 1 is incorporated into strand #1. Strand #1 has a region which is complementary to strand #2. Primer 2 is downstream from primer 1, binds to strand #1, and is incorporated into the strand antisense to strand #1. Primer 3 is incorporated into strand #3. Strand #3 differs in part from strand #1. Primer 4, which is identical to primer 2 in this illustration, is down- stream from primer 3, binds to strand #3, and is incorpo¬ rated into the strand antisense to strand #3. Following annealing of strand #1 to the strand antisense to strand #3, there exists a region of strand #1 with complementari- ly to strand #2. a,b,c and d are the products of PCR amplification. (Primer 2 and primer 4 are usually the same primer. )

Figure 8 illustrates the joining of the single ends of two double-stranded DNA molecules. Primer 1 binds to strand #1 of DNA segment #1. The strand into which primer 1 is incorporated is at least partially complemen¬ tary to the strand into which primer 3 is incorporated. Primer 2 binds to strand #2 of DNA segment #1 and is downstream from primer 1. Primer 5 binds to strand #1 of DNA segment #1, and is incorporated into a strand which has homology to the strand into which primer 1 is incorpo¬ rated, but which also differs, at least in part, from the strand into which primer 1 is incorporated. (In the figure, the 5' terminus of primer 5 is downstream from the 5' terminus of primer 1). Primer 3 binds to strand #2 of DNA segment #2. Primer 4 binds to strand #1 of DNA segment #2 and is downstream from primer 3. Primer 6 binds to strand #2 of DNA segment #2, and is incorporated into a strand which has homology to the strand into which primer 3 is incorporated. The strand into which primer 6 is incorporated also differs, at least in part, from the strand into which primer 3 is incorporated. a,b,c, and d aro the products of PCR amplification. (In the figure, the 5' terminus of primer 6 is downstream from the 5' terminus of primer 3).

Figure 9 illustrates the circularization of a DNA segment. Primer 1 binds to strand #1 of DNA segment #1. The strand into which primer 1 is incorporated is at least partially complementary to the strand into which primer 3 is incorporated. Primer 3 binds to strand #2 of DNA segment #1. Primer 2 binds to strand #2 of DNA segment #1. It is downstream from primer 1 and is incorporated into a strand which has homology to the strand into which primer 3 is incorporated, but differs, at least partially, from the strand into which primer 3 is incorporated. (In this illustration, the 5" terminus of primer 2 is down¬ stream from the 5' terminus of primer 3) . Primer 4 binds to strand #1 of DNA segment #1. it is downstream from primer 3 and is incorporated into a strand which has homology to the strand into which primer 1 is incorporat¬ ed, but differs, at least partially from the strand into which primer 1 is incorporated. a, b, c and d are the products of PCR amplification. (In this illustration, the 5' terminus of primer 5 is downstream from the 5' terminus of primer 1) .

Figure 10 illustrates the joining of both ends of two DNA segments to one another. Primer 1 binds to strand #1 of DNA segment #1. The strand into which primer 1 is incorporated is at least partially complementary to the strand into which primer 5 is incorporated. Primer 3 binds to strand #2 of DNA segment #1, and the strand into which it is incorporated is, at least partially, co ple- mentary to the strand into which primer 7 is incorporated. Primer 5 binds to strand #1 of DNA segment #2. Primer 7 binds to strand #2 of DNA segment #2. Primer 2 binds to strand #2 of DNA segment #1. It is downstream from primer 1 and is incorporated into a strand which has homology to the strand into which primer 3 is incorporated. The strand into which primer 2 is incorporated also differs, at least partially, from the strand into which primer ;3 is incorporated. (In this illustration, the 5' terminus of primer 2 is downstream from the 5-' terminus of primer 3) . Primer 4 binds to strand #1 of DNA segment #1. It is downstream from primer 3 and it is incorporated into a strand which has homology to the strand into which primer 1 is incorporated. The strand into which primer 4 is incorporated also differs, at least partially, from the strand into which primer 1 is incorporated. (In this illustration, the 5' terminus of primer 4 is downstream from the 5' terminus of primer 1). Primer 6 binds to strand #2 of DNA segment #2. It is downstream from primer 5, and it is incorporated into a strand which has homology to the strand into which primer 7 is incorporated. The strand into which primer 6 is incorporated also differs, at least partially, from the strand into which primer 7 is incorporated. (In this illustration, the 5' terminus of primer 6 is downstream from the 5' terminus of primer 7) . Primer 8 binds to strand #1 of DNA segment #2. It is downstream from primer 7 and it is incorporated into a strand which has homology to the strand into which primer 5 is incorporated. The strand into which primer 8 is incorporated also differs, at least partially, from the strand into which primer 5 is incorporated. a, b, c and d are the products of PCR amplification. (In this illus¬ tration, the 5' terminus of primer 8 is downstream from the 5' terminus of primer 5).

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for generating DNA joints. Such a method can be employed for the rapid generation of a site-specific mutant and for DNA recombination.

In the present method, separate PCR reactions or amplifications are utilized to generate products that when combined, denatured and reannealed, form double-stranded DNA with discrete, cohesive single-stranded ends, in addition to the original blunt ended products. The generation of these cohesive ends of DNA permits the formation of precise, directional DNA joints without dependence on enzyme restri 'ion sites. Ftrthermore, the placement of coheε ^re .sing! >' strands on- the ends of a DNA molecule permits a DNA strand or strands to be annealed to form circular DNA. This method is designated as recombi¬ nant circle PCR (RCPCR) . These recombinant circles of DNA can be directly transfected into various host cells, for example, E. coli cells. There is no need for phosphoryla- tion of primers, restriction enzyme digestion, or a ligation step. Furthermore, each manipulation, from setting up for the DNA amplification to transfection into the microorganism of choice can be accomplished in only one day. Thus, the method of generating recombinant circles upon reannealing PCR generated products greatly simplifies DNA mutagenesis, recombination, and cloning since no enzymatic step is needed apart from the PCR amplifications.

The use of the present invention for site-specific mutagenesis and for recombination is described in detail below.

Site-Specific Mutagenesis In the first embodiment of the invention, a site- specific mutation can be formed by the generation of nicked or gapped recombinant circles, as illustrated in Figure 1. An uncut plasmid with the insert of interest is simultaneously amplified and mutated in two separate PCR reactions. In each of these two reactions, the plasmid is amplified, and the identical base pair (bp) or region is mutated. The primers are designed so that the "break point" of the plasmid in one PCR product is different from the "break point" of the plasmid in the other PCR product. This "break point" is the location between the two 5' ends of the amplifying primers in any one reaction. Due to the placement of these break points, if one strand from one PCR product anneals to its complementary strand from the other PCR product, a discrete single-stranded segment will extend from each end. These segments or single strands wil^"1 be complementary to each other, permitting the formation of circles of DNA without restriction enzyme digestion or ligation _in vitro. , ,*

^■ The DNA can then be transfected into appropriate host cells, for example, E. coli cells, and a determina¬ tion- can be made as to whether the desired mutation is present in the transfected cells.

The ability to make site-specific mutations by using RCPCR to generate gapped circles, as opposed to nicked circles, offers several advantages. It broadens the choice of potential primer annealing sites, and lessens the cost of making multiple mutations by allowing the use of two conserved "outside" primers (primers 2 and 4 in Figure 1; the primers furthest outside the mutagene¬ sis site) while using several different sets of "inner primers" (primers 1 and 3 in Figure 1) that not only contain different mutations, but also span different areas. Therefore, each new mutation in a region would require one to make only two new primers.

Furthermore, there is considerable flexibility in the location of the mismatched nucleotide in a primer under the amplification conditions of the present inven- tion. This is true since sufficient amplification occurs with a mismatch, near the 3' terminus of the mutating primers, for recombinant circles to be formed in suffi¬ cient quantity such that they can be transfected into appropriate host cells. By modification of the primers incorporated into the ends of the PCR generated products, it is clear that mutagenesis by RCPCR can be used to generate insertion (see Figure 3) and deletions (see Figure 4) .

It should also be noted that if primers 2 or 4 contained a mutation in addition to primers 1 and 3, two distal regions could be mutated at once. The mutation caused by primer 2 or 4 would result in approximately 50% of the clones containing both mutations.

RCPCR can also be utilized for DNA recombination, as described below.

DNA Recombination The second embodiment of the invention, illustrat¬ ed by Figure 2, involves a strategy for amplifying an i. sert from one- * -asmid, and inserting it into a differ- ent, recipient plasmid in a specific location with a defined orientation. This particular DNA recombinational event can be viewed as a directional subcloning. The insert is amplified with primers 1 and 2. The insert is both amplified and modified with primers 3 and 4. The 5' ends of primers 3 and 4 contain regions homologous to specific regions in the recipient plasmid. Therefore, if one reanneals the products of these two PCR reactions, products are formed with single-stranded ends that can anneal to recipient plasmid sequences. The recipient plasmid is amplified with primers 5 and 6, and in a separate reaction with primers 7 and 8. These primers are designed so that when these products are combined, dena- tured, and reannealed, products are formed with single- stranded ends that will anneal to the single-stranded sequences protruding from the insert sequence. The single-stranded ends could extend from one or both strands of the insert or recipient plasmid, as long as the single- stranded ends extending from the insert will anneal to the single-stranded ends extending from the recipient plasmid. In the protocol, the products from all 4 PCR amplifica¬ tions are combined, and the denaturation and annealing occurs in a single reaction mix. It should be clearly noted that the procedure outlined above provides a general strategy for DNA recom¬ bination. It can be used to transfer one DNA molecule into any point of another DNA molecule, with or without replacing a segment in the recipient DNA (see Figure 6) . Therefore, it could be used, for example, to fuse genes or to create hybrid promoters.

In a third embodiment of the present invention, insertional mutagenesis can be achieved by the generation of recombinant circles of DNA (see Figure 3) . Insertional mutagenesis can also be accomplished by placing an excess number of nucleotides in the ^mismatched site of primers 1 anr- 3 of Figure 1.

Furthermore, in a fourth embodiment, mutagenesis can be accomplished by inserting a synthetic oligonucleo- tide patch into 'a recipient plasmid. The generation of recombinant - circles would occur by using a synthetic oligonucleotide with cohesive single-stranded ends in the recipient plasmid. The synthetic oligonucleotide with cohesive single-stranded ends takes the place of the insert with the cohesive single-stranded ends in Figure 2.

In a fifth embodiment, deletional mutagenesis can be achieved by the generation of recombinant circles of

DNA (see Figure 4) . Deletional mutagenesis could also be accomplished by leaving out nucleotides in the mismatch site of primers 1 and 3 of Figure 1.

In a sixth embodiment, site-specific mutagenesis of a very large vector/insert can be obtained by the generation of recombinant circles of DNA (see Figure 5) . This embodiment differs from regular site-specific muta¬ genesis in that the plasmid is broken apart by a total of four PCR amplifications, so that the length of each amplification product can be reduced. For each of the PCR amplifications, there is a large region of the vector in which restriction enzyme digestion can occur without adversely effecting the amplification process. This linearization of the template can increase the total length and amount of the product that can be formed during a given PCR reaction.

In a seventh embodiment of the present invention, an insert is substituted for a sequence of DNA in a recipient plasmid using RCPCR (see Figure 6) . Primer 1 and primer 2 contain only sequences contained in the ends of the insert. Primers 3 and 4 have 3' ends that are identical to primers 1 and 2 and 5' ends that are homolo¬ gous to the region in which the insertion is to take place on the recipient plasmid. Primers 5 and 6 amplify the recipient plasmid minus the region which is to be removed during the substitution, and contain the regions comple¬ mentary to the 5' regions, of primers 3 and 4. Primers 7 and 8 ampli y a region cf the recipient plasmid which is ^■ ternal to tht 5' ends of- primers 5 and 6.

In an eighth embodiment-of the present invention, the single ends of two., doable-stranded molecules can be joined (see Figure j).-

In a ninth embodiment of the present invention, a DNA segment can be circularized (see Figure 9) . Basical¬ ly, if the primers are phosphorylated at the 5' termini, and the recombinant circles are ligated, closed circular DNA can be generated jLn vitro, using any RCPCR technique. In this embodiment, as opposed to the eighth embodiment (see Figure 8), only four primers are required because primer 2 fulfills the functions on primers 2 and 6 of the eighth embodiment. Primer 4 fulfills the functions of primers 4 and 5 of the eighth embodiment.

In a tenth embodiment of the present invention, both ends of two DNA segments can be joined to one another (see Figure 10).

In an eleventh embodiment of the present inven¬ tion, multiple nucleotide substitution mutagenesis can be achieved. Specifically, this can be accomplished by using a mismatched region spanning several nucleotides in Example 1. This type of mutagenesis could also be achieved by modifying the placement of primers 1 and 3 in the insertional mutagenesis protocol shown in Figure 3. Specifically, the insertional sequence is substituted for a specific sequence of DNA.

A twelfth embodiment of the invention involves gene fusion. This can be accomplished by the DNA recombi¬ nation techniques illustrated in Figures 2 and 6 wherein the insert from the donor plasmid is one gene, and a region immediately adjacent to the subcloning site in the recipient plasmid is another gene.

In a thirteenth embodiment of the invention, vector reconstruction can be accomplished by recombination of several PCR amplified segments into a vector. Since the location of the DNA joints is not restricted by restriction enzyme sites or the use of linker or adaptors, with this procedure, and the generation of transfectable circles can take place without any enzymatic step beyond the original PCR amplifications, considerable flexibility and speed is; allowed in generating novel vectors, hybrid promoters, or any DNA construction. For example, the orientation of a specific segment of DNA could be reversed to test the effect of orientation on function. Further¬ more, the above type of vector reconstruction will be of particular value when convenient restriction enzyme sites are not available.

In a fourteenth embodiment of the present inven¬ tion, the- -components described in the site-specific mutagenesis protocol can be placed in a kit. This kit can be used for site-specific mutagenesis of different inserts if the components of embodiments 1, 2, 3, 6 or 7 are included, except for the mutating primers which have to be individually designed for each insert. The kit consists of a double-stranded plasmid, and in the simplest case, two primers possessing 5' ends. These 5' ends bracket the insertion site for the insert which is to be mutated. Each of these two primers anneal to separate strands of the double-stranded plasmid, with their 3' ends pointing away from each other. These primers correspond to primers 2 and 4 in embodiments 1, 3, and 5. The primers can also correspond to primers 2 , 3, 4, 5, 7, 8 in embodiment 6, or primers 3 and 7 with strands generated by primers 2 and 4 and by primers 5 and 8. Since RCPCR is effective follow¬ ing the generation of gapped circles, the size of the insert can vary. This kit can also include annealing buffer, and even highly competent frozen cells to facili¬ tate highly efficient transfection of the recombinant circles. Furthermore, to increase transfection efficien¬ cy, the primers and/or DNA strands placed in the kit can be phosphorylated, and ligation buffer and ligase can be included.

In a fifteenth embodiment of the invention, the recipient vector described in Example 2 can be placed in a kit to facilitate subcloning of PCR generated linear products. The recipient vector is linear with cohesive single-stranded .ends. The fragment that is to be sub- cloned is* amplified in two separate PCR-reactions, such that when they are combined, denatured, and reannealed, double-stranded products are formed with single-stranded ends that will anneal to the single-stranded ends protrud¬ ing from the recipient vector in the kit. These single- stranded cohesive ends can vary from 10 to several 100 base pairs. The 5' single-stranded end(s) can be phos¬ phorylated at the 5' terminus, in which case the kit can include ligase (and ligase buffer), but this would not be necessary. The kit can also include annealing buffer, and highly competent frozen cells to facilitate highly effi¬ cient transfection of the recombinant circles.

In a sixteenth embodiment of the present method, wherein site-specific mutagenesis is involved, one of the plasmids in which the Pstl site had been eliminated underwent RCPCR using primers designed to re-establish the Pstl site (see Example 1). In this reaction the mis¬ matched nucleotide was only 4 and 3 nucleotides from the 3' and in primers 1 and 3, respectively. The number of clones generated was diminished 50% by the use of these primers. Plasmids from 10 of 10 of the resulting colonies had an intact Pstl site.

In a seventeenth embodiment of the present inven¬ tion, one end of a double-stranded DNA molecule can be joined to a single-stranded DNA molecule (see Figure 7).

RCPCR can also be utilized to manufacture circles with protruding single-stranded ends, which can be used for hybridization or for strand invasion in homologous recombination. Furthermore, the strategy of generating cohesive single-stranded ends upon reannealing PCR ampli¬ fied and/or modified products can also be used to manufac¬ ture complex shapes of DNA molecules, that may be used alone or following fusion to proteins or other biological moieties. The DNA can function as structural domains, catalytic domains, or as hybridization probes to target DNA or other attached biological moieties to specific DNA or RNA sites. DNA can be manipulated with a latitude that may not have a biologic parallel.

The method of the present invention can be illus- trated by the following non-limiting examples.

EXAMPLE 1 Site-Specific Mutagenesis Initially, the use of RCPCR for site-specific mutagenesis was tested by eliminating the Pstl site in the lacZ' operon of pUC19 by the generation of nicked circles. This was accomplished by substituting a T for the G at nucleotide #437.

In one experiment, 15 of the 18 resulting plasmids lacked the Pstl site, and in another experiment, 10 of 10 resulting plasmids lacked the Pstl site. Sequencing of 550 nucleotides, containing the lacZ¹ operon, revealed only the desired mutation in 6 of 6 plasmids. Gapped circles of DNA were made as follows:

Formation of Gapped Circles of DNA:

Primers 2 and 4 were each synthesized with their 5' terminus 10 bp lateral (see Figure 1) to the primers used to make the nicked circles. This resulted in recom- binant circles containing two 10 bp gaps. Plasmids from 12 to 12 resulting colonies lacked the Pstl site. 200 nucleotides surrounding the Pstl site, and including each gap, were sequenced in 4 of the plasmids, revealing only the desired mutation, there was no decrease in the number of clones obtained.

Initially, the plasmids were prepared as followed: Preparation and Purification or Plasmids:

Plasmids for PCR or DNA sequencing were prepared by Triton X-100/lysozyme lysis and cesium chloride banding (Davis et al, Basic Methods in Molecule Biology, pp. 93-98 (1986)). The plasmids utilized for PCR were not restric¬ tion enzyme digested prior to amplification. Oligonucleo¬ tide primers were prepared on an Applied Biosystems DNA synthesizer in the trityl on mode. They were purified on Applied Biosystems purification cartridges, and were not ^***• phosphorylated.

Fourteen cycles of amplification were then carried out as follows: PCR Amplification' PCR amplification was performed with Thermus aquaticus DNA polymerase (Taq polymerase) using a Perkin- Elmer Cetus Thermal Cycler. Amplifications were in Taq buffer with 25 p oles each primer, 2 ng template, and 1.25 U Taq polymerase (Perkin Elmer Cetus), in 50 μl. Reac- tants underwent initial denaturation (94°C x 1 min) , 14 amplification cycles (94°C x 30 sec, 50°C x 30 sec, and 72°C x 2.5 min for mutagenesis or 72°C x 3 min for direc¬ tional subcloning) and a final extension (72°C x 7 min) . Each product was readily visualized following agarose gel electrophoresis. These products were puri¬ fied, and then combined, denatured and reannealed as indicated below: Purification of PCR Products:

Each PCR product underwent electrophoresis over a 1% agarose gel. The products were stained with ethidium bromide, visualized with UV light, cut out, and removed from the gel with GENECLEAN (manufactured by BIO 101). Denaturation and Reannealing:

The purified PCR products were combined in 50 μl annealing buffer (0.1M NaCl in lOmM Tris-HCL pH8/lmM EDTA) in the mutagenesis protocol, and in 100 μl annealing buffer in the directional subcloning protocol. They were denatured at 94°C x 3 min, reannealed at 50°C x 2 hours, and then placed at room temperature.

The formation of recombinant circles could be visualized by agarose gel electrophoresis, in addition to linear DNA. The annealing mix was transfected into E_j_ coli and the colonies were screened as indicated below: Transfection Into E. Coli:

1-5 μl of the annealed mix, without gel purifica¬ tion of the recombinant circles, was transfected directly into HB101 E. coli. (BRL, Life Technologies) following the manufacturer's protocol. Colonies were not color selected for elimination of lacZ' complementation (Maniatis et al, Molecular Cloning: A Laboratory Manual, pp. 51-52 (1982)). Transfection yielded 400-1000 ciones/ng recombinant circle DNA in the mutagenesis protocol, and 60 clones with the insert in the recipient plasmid/ng recombinant circle DNA" in the subcloning protocol. Screening of Colonies:

Colonies were grown overnight in LB broth. Minipreps were prepared by the alkaline lysis method (Maniatis et al, Molecular Cloning: A Laboratory Manual, pp. 368-69 (1982)). Restriction enzyme analysis was performed using conventional methods. DNA Sequencing:

Plasmids were sequenced by the dideoxy method (Sanger et al, Proc. Natl. Acad. Sci. USA 74:5 463-67 (1977)) . EXAMPLE 2

Directional Subcloning A 707 bp insert was amplified from the polylinker site of pUC19 and placed in the opposite orientation in the polylinker site of BLUESCRIPT (manufactured by Stratagene). Following transfection, 3 out of 10 colonies tested contained inserts, and all of the inserts were in BLUESCRIPT and were in the correct orientation. Sequenc¬ ing of 150 bp surrounding the two junctional regions in each of the 3 clones (total 900 nucleotides) revealed no aberrant sequences within 50 bp of each junctional region. The insertion of a T between two Cs was detected in one of the inserts, indicating a 0.11% misincorporation error frequency in the construct. (See Example 1 for a discus¬ sion of amplification, transfection, sequencing, as well as the primary steps utilized in such a procedure, for example, preparation and purification of the plasmids.

Claims

WHAT IS CLAIMED:

1. A method for synthesis of a double-stranded DNA molecule using the polymerase chain reaction process, wherein the nucleotide sequence of the first single-strand of said double-stranded DNA molecule has at least one difference from the complement of the nucleotide sequence of the second single-strand of said double-stranded DNA molecule, comprising the steps of:

(i) providing a single-stranded DNA tem- plate, any desired segment of the nucleotide sequence of said template being used for synthesis of said first strand of said double-stranded DNA molecule; and

(ii) providing a first polynucleotide primer, at least a portion of the nucleotide sequence of which is complementary to said template starting at the nucleotide that defines the 3' end of said desired segment and extending in the 5' direction of said template; and

(iii) providing a second polynucleotide primer, at least a portion of the nucleotide sequence of which is complementary to said template, starting at said 3' end of said desired segment or starting 5' to said 3' end, and extending in the 5' direction of said template, wherein there is at least one difference between said nucleotide sequence of said first primer and said nucleotide sequence of said second primer, said difference comprising a substitution, deletion, insertion or extension of one or more nucleotides, provided that, if both of said first and -second primers are complementary to said template starting at the same nucleotide on said 3' end of said desired segment, and if said difference in said first and second primers is entirely 5' to said 3' end of said desired segment, then said difference does not consist entirely of an extension of the sequence that is complementary to said template in one of said primers with respect to the other; and

(iv) providing a third polynucleotide primer, at least a portion of the nucleotide sequence of which is homologous to said single-stranded template starting at the nucleotide that defines the 5' end of said desired segment and extending in the 3' direction of said template; and (v) synthesizing double-stranded DNA molecules using said template and said primers and the polymerase chain reaction process; and

(vi) separating the primer extension prod¬ ucts resulting from step (v) from the strands to which they are hybridized so that single-stranded extension products are formed; and

(vii) reannealing the single strands result¬ ing from step (vi).

2. The method according to claim 1, for produc- ing a circular double-stranded DNA molecule, in which steps (v), (vi) and (vii) further comprise the steps of: in a first container means containing a first aliquot of a DNA template;

(i) treating a first strand of said tem- plate with a first primer having a sequence that is suffi¬ ciently complementary to a first region of said first strand to hybridize therewith under conditions such that hybridization between said first primer and said first region of said first strand occurs; (ii) treating a second strand of said template with a second primer having a sequence that is sufficiently complementary to a first region of said second strand to hybridize therewith under conditions such that^* hybridization between said second primer - and said first region of said second strand occurs;

(iii) treating the products of steps (i) and (ii) using denaturation and primer reannealing such that an extension product of each of said first and second primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said first and second primers, when separated from the strand to which it is hybridized, can serve as a template for synthesis of the extension product of the other of said first and second primers;

(v) treating the single-stranded extension products from step (iv) with said first and second primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (iv) as a template, whereby a first double-stranded product is formed; in a second container means containing a second aliquot of said template; (vi) treating said second strand of said template with a third primer having a sequence that is sufficiently complementary to a second region of said second strand to hybridize therewith under conditions such that hybridization between said third primer and said second region of said second strand occurs, said third primer having a sequence that is sufficiently complementa¬ ry to the sequence of said first primer for hybridization to occur;

(vii) treating said first strand of said plasmid with a fourth primer having a sequence that is sufficiently complementary to a second region of said first strand to hybridize therewith under conditions such that hybridization between said second region of said first strand and said fourth primer occurs, said second region of said first strand being distinct from said first region of said second strand;

(viii) treating the products of steps (vi) and (vii) under conditions such that an extension product of each of said third and fourth primers is synthesized which is complementary to each of said second and first strands, respectively, such that the extension product synthesized from one of said third and fourth primers, when separated from the strand to which it is hybridized. can serve as a template for synthesis of the extension product of the other of said third and fourth primers;

(ix) separating the primer extension prod¬ ucts resulting from step (viii) from the strand to which it is hybridized so that single-stranded extension prod¬ ucts are formed; and

(xi) combining the first and second double- stranded products resulting from steps (v) and (x); (^χϋ) separating the strands of said first and second double-stranded products resulting from steps (v) and (x) ;

(xiii) reannealing the single strands result¬ ing from step (xii); and (^χiv) selecting circular DNA molecules comprising the extension products of said first, second, third and fourth primers.

3. The method according to claim 2 wherein said first and third primers contain at least one base that is not complementary to said first region of said first strand of said template and said second region of said second strand of said template, respectively, and wherein said at least one base of said first primer is complemen¬ tary with said at leaε one base of said third primer.

4. The method according to claim 1, for trans¬ ferring an insert contained in a donor plasmid to a recipient plasmid in which (v), (vi) and (vii) further comprise the steps of: in a first container means containing an aliquot of donor plasmid containing an insert:

(i) treating a first strand of said donor plasmid with a first primer having a sequence that is sufficiently complementary to a region of said first strand to hybridize therewith under conditions such that hybridization between said first primer and said region of said first strand occurs, wherein the first primer con¬ tains only a sequence present in one end of the insert; (ϋ) treating a second strand of said donor plasmid with a second primer having a sequence that is sufficiently complementary to a region of said second strand to hybridize therewith under conditions such that hybridization between said second primer and said region of said second strand occurs, wherein the second primer contains only a sequence present in the other end of the insert; in a second container means containing an aliquot of a donor plasmid containing an insert: (ϋi) treating said first strand of said donor plasmid with a third primer having a sequence that is sufficiently complementary to said region of said first strand to hybridize therewith under conditions such that hybridization between said third primer and said region of said first strand occurs, wherein the 3' end of said third primer is identical to primer 1 and has several base pairs on its 5' end which are homologous to a region on the recipient plasmid of step (xi);

(iv) treating said second strand of said donor plasmid with a fourth primer having a sequence that is sufficiently complementary to said region.,, of said second strand to hybridize therewith under conditions such that hybridization between said fourth ,primer and said region of said second strand occurs, wherein the 3' end of said fourth primer is identical to primer 2 and has several base pairs on its 5' end which are homologous to a region on the recipient plasmid of step (xi);

(vi) separating the primer extension prod- ucts resulting from step (v) from the strands to which they are hybridized so that single-stranded extension products are formed; and

(vii) treating the single-stranded extension products from step (vi) with said first and second primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (vi) as a template, whereby a first double-stranded product is formed;

(viii) treating the products of steps (iii) and (iv) under conditions such that an extension product of each of said third and fourth primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said third and fourth primers, when separated from the strand to which it is hybridized, can serve as a template for synthesis of the extension product of the other of said third and fourth primers;

(ix) separating the primer extension pro¬ ducts resulting from step (viii) from the strands to which they are hybridized so that single-stranded extension products are formed; and π.

(x) treating the single-stranded extension products from step (ix, .-"1th said third and fourth primers under conditions such tnat a primer extension product is synthesized using each of said single-stranded extension products of step (ix) as a template, whereby a first double-stranded product is formed; in a third container means containing an aliquot of a recipient plasmid: (^χi) treating the first strand of said recipient plasmid with a sixth primer having a sequence that is sufficiently complementary to a first region of said first strand to hybridize therewith under conditions such that hybridization between said sixth primer and said first region of said first strand occurs, said sixth primer containing the region complementary to the 5' regions of primer 3; ( ϋ) treating the second strand of said recipient plasmid with a fifth primer having a sequence that is sufficiently complementary to a first region of said second strand to hybridize therewith under conditions such that hybridization between said first region of said second strand and said fifth primer occurs, said fifth primer containing the region complementary to the 5' region of primer 4; in a fourth container means containing an aliquot of a recipient plasmid; (xiii) treating the first strand of said recipient plasmid with an eighth primer having a sequence that is sufficiently complementary to a second region of said first strand to hybridize therewith under conditions such that hybridization between said eighth primer and said second region of said first strand occurs, said second region of said first strand being distinct from said first region of said first strand;

(xiv) treating the second strand of said recipient plasmid with a seventh primer that is suffi- ciently complementary to a second region of said second strand to hybridize therewith under conditions such that hybridization between said seventh primer and s-riΛiisecond region of said second strand occurs, said second region of said second strand being distinct from said first region of said second strand;

(xv) treating the products of steps (xi) and (xii) under conditions such that an extension product of each of said sixth and fifth primers is synthesized which is complementary to each of said first and second strands, respectively, such that the extension product synthesized from one of said sixth and fifth primers, when separated from the strand to which it is hybridized, can serve as a template for the synthesis of the extension product of the other said sixth and fifth primers;

(xvi) separating the primer extension prod¬ ucts resulting from step (xv) from the strands to which they are hybridized so that single-stranded extension products are formed; and

(xvii) treating the single-stranded extension products from step (xvi) with said sixth and fifth primers under conditions such that a primer extension product is synthesized using each of said single-stranded extension products of step (xvi) as a template, whereby a first double-stranded product is formed;

(xix) separating the primer extension prod¬ ucts resulting from step (xviii) from the strands to which they are hybridized so that single-stranded extension products are formed; and ( ^χ) treating the single-stranded extension products from step (xix) with said eighth and seventh prxmers under conditions such that a primer extension product i synthesized using each of said single-stranded exte: sion products of step (xix) as a -template, whereby a / first double-stranded product is formed;

(xxi) combining the double-stranded products resulting from steps (vii), (x), (xvii) and (xx) ;

(xxii) separating the strands of said double- stranded products resulting from steps (vii), (x), (xvii) and (xx) ; and

(xxiii) reannealing the single-strands resulting from step (xxii).

5. A method of producing a site-specific uta- tion in an autonomously replicating double-stranded circular DNA comprising the steps of: in a first container means containing a first aliquot of said circular DNA; (i) contacting said circular DNA with a first primer and second primer, wherein said first primer is complementary to a region of the first strand of said circular DNA, except that it contains a mismatched base at the site at which mutagenesis is to end of said first primer, said second primer is complementary to the second strand of said circular DNA, wherein the 5' ends of said first and second primers bind at contiguous but non-over lapping regions of said first and second strands, and wherein said primers are in substantial molar excess over said strands,

(ii) producing non-circular copies of said second strand containing said mismatched based and non- circular copies of said first strand by means of the polymerase chain reaction, in a second container means containing a second aliquot of said circular DNA:

(iii) contacting said circular DNA with a third primer and fourth primer, wherein said third primer is complementary to a region of said second strand, except that it contains a mismatched base at the site at which mutagenesis is to occur, . said mismatched base being at least 8 bases from the 5' end of said third primer, said third primer containing a sequence including said mis¬ matched base which is complementary to a fourth primer which is complementairy to said first strand, wherein the 5" ends of said third and fourth primers bind at contigu¬ ous but non-overlapping regions of said first and second strands, and wherein said primers are in substantial molar excess over said strands, (iv) producing non-circular copies of said first strand containing said mismatched base and non- circular copies of said second strand by means of the polymerase chain reaction, (v) combining the products of steps (ii) and (iv),

(vi) subjecting the mixture of step (v) to denaturation conditions to produce single-stranded DNA, ( ϋ) subjecting the mixture of step (vi) to annealing conditions to produce double-stranded DNA, and

(viii) isolating the circular DNA containing the mismatched bases.

6. The method of claim 5 wherein said autono- mously replicating double-stranded circular DNA is a plasmid.